With plenty of debate out there about the sustainability of the current rally, we thought it might be interesting to approach performance data for the S&P 500 in a different way and see how the current rally stacks up against historical data on a rolling 1-quarter, 2-quarter, 3-quarter, and 4-quarter basis in terms of consecutive quarters of positive performance. Ignoring the magnitude of performance gains, several interesting observations stick out. As usual, let’s move to the data.
There are a few notes regarding methodology. First, the period covered stretches back to 1950, a total of 255 quarterly observations. Second, the rolling period returns represent simple, nominal returns and therefore don’t account for dividends or inflation. Finally, when looking at the rolling 3 and 4 rolling quarter periods, a single negative quarterly result did not break a streak if the loss didn’t exceed 1%. There were a few instances in the middle of longer periods, including the current period, where markets took a slight breather, but resumed climbs for several more quarters. We felt this kept with the spirit of the discussion, i.e. trying to understand the persistence of rallies and positive market performance from a calendar perspective.
As of this morning in the S&P 500, the market is marking its 1st consecutive quarter of 1-qtr positive returns (Q4:2012 was negative), 5th consecutive quarter of positive rolling-2-quarter returns, the 5th straight quarter of positive rolling-3-quarter returns, and 14th quarter of positive rolling-4-quarter returns (the exception mentioned above applies to the quarter ending 9/30/11, which shows a -0.86% year over year return. Nonetheless, the market has been in a relatively consistent up-streak since the 2009 lows). Since 1950, the average quarterly streak for single quarter positive performance is 3.11 quarters. The average streak for rolling-2-month positive performance without an interruption is 5.18 quarters. The average streak for 3-rolling-quarter positive performance is 9.58 quarters. The average for a trailing 4-quarter period is 11.50 quarters. As the number of rolling periods decreases, the numbers become volatile. Hence the average length of a streak increases as we move from 1 performance period to a rolling-4 calculation.
 |
*Exception described above in methodology applies.
|
As with other data we’ve looked at over the past several weeks, the overall market environment makes a difference when looking at these numbers, though only at the one, two, and three quarter rolling points. During secular bear markets, still technically our current situation, the average positive performance streaks are as follows: 2.43 quarters for single quarter performance in isolation; 4 quarters for 2-quarter rolling performance; 8.57 for 3-quarter rolling performance; and 12.2 for 4-quarter rolling performance. For secular bull markets: 3.63 for single quarters; 6.05 for rolling 2-quarters; 10.17 for rolling 3-quarter; and 11.18 for rolling 4-quarter. It’s interesting that the positive streaks for rolling 4-quarter performance are actually longer during secular bear markets than during secular bull markets. We do know from our post several weeks ago, however, that overall rally performance during secular bulls is much stronger.
The current value of 5.68% rolling two-quarter performance comes in above the average of 4.52%, but is exceeded by 116 of 255 observations. Rolling 3-quarter performance of 11.77% comes in above the average of 6.77%, and is exceeded by 90 observations. Rolling 4-qtr performance of 8.02% is below the long-term average of 8.99%, but exceeded by 136 observations. Of the 90 observations ahead of the rolling-3-quarter number, there’ve only been four instances when the next calendar observation turned negative: 9/30/2011, 6/30/2010, 9/30/2000, and 12/31/1987. Three of the four instances involved major short-term external events, or short-term instances, that shocked the markets. In 2010 and 2011, flare ups in European credit markets caused performance to decline sharply before resuming. In 1987, several factors combined to cause the harrowing October ’87 stock market crash. Of the 136 rolling-4-quarter observations ahead of the current performance-wise, only seven times has the following quarterly observation turned negative: 9/30/2011, 12/31/2000, 9/28/1990, 9/30/1981, 3/31/1977, 6/30/1969, and 3/31/1960. Therefore, in most cases strong quarterly performance, at least when looking at several rolling periods, tends to morph into slower positive performance gains before turning negative rather than experiencing a sharp or immediate reversal.
Ok, there are a lot of numbers here, so what can we take away. Over the past 60+ years, there’s generally been strong persistence on a calendar basis when it comes to US markets generating positive performance numbers. Although the numbers differ somewhat, the persistence in calendar terms holds up reasonably well during secular bear markets, though we know the magnitude of secular bear market rallies overall in terms of actual performance significantly lags the performance of rallies during secular bull market periods. The numbers may give some additional comfort to traders out there using momentum as a factor in investment decision making. As for the current situation, things are a bit muddier. In calendar terms, the rally is long in the tooth when considering the fact that there have been 14 quarters of positive year over year performance (which of course, includes the one slight hiccup we discussed in the exceptions near the top). The three month rolling number streak is relatively young by historical standards, though. As we’ve observed in other blog posts, valuation for the US is stretched by historical standards, yet the magnitude of the current rally off the last major correction is below average compared to other cyclical bull runs. Since longer-term valuations have never reached levels associated with secular bear market ending points and are actually much closer to points associated with secular bull ending points, we think the current rally, at least in the US, represents a rally within a secular bear, not the kickoff to a new multi-year positive streak as witnessed in the ‘80s and ‘90s. Market sentiment has begun to creep higher, but at this point hasn’t reached extreme positive levels across the board, a good thing for the bulls. Taking all into consideration, we believe the current rally could have legs for a few more quarters in the US, but that caution could be warranted, perhaps sometime in late 2013. This cyclical bull market rally may be in the later innings, but as the numbers above suggest, not to mention past episodes with above average valuations, rallies can last much longer than many people imagine.
